Galactosemia due to deficiency of galactose-1 phosphate uridyl transferase activity (GALT) affects 1 in every 50,000 newborn infants. Despite newborn screening programs in most states and readily available treatment via dietary galactose restriction, the prognosis for normal physical and mental development is less optimistic than originally thought. Even with the best current therapy, many affected children are developmentally retarded and all affected females have primary ovarian failure. The first objective of this proposal is to test the hypothesis that clinical variation is related to genetic heterogeneity; knowledge of the genotype might lead to customized therapy or at least a more accurate prognosis. Because questions about developmental and tissue specific effects of GALT deficiency cannot be adequately studied in humans, the second objective will be to develop an animal model for GALT deficiency, using gene targeting in embryonic stem cells. The specific aims of this proposal are: 1. Define the genotype in transferase deficiency galactosemia and relate the finding to clinical phenotype. To do this we will: a) characterize mutations in the GALT genes of patients with altered GALT activity by sequence analysis b) determine the effect of specific mutations of GALT activity by transfecting GALT negative eucaryotic cells with mutant cDNA construction. c) correlate genotype findings with patient phenotype in a group of patients with classical galactosemia. 2. Develop a mouse model for transferase deficiency galactosemia. To do this we will: a) isolate and characterize the mouse GALT gene b) develop a gene targeting construction for electroporation into mouse embryonic stem cells; identify targeted stem cell colonies c) breed a homozygous GALT negative (GALT KO) mouse and characterize the effects of galactose toxicity on fetal development and tissue specific pathology in these animals.